Font Size: a A A

Research And Implementation Of Academic Conferences Recommendation System Based On Hybrid Recommendation Strategy

Posted on:2020-06-04Degree:MasterType:Thesis
Country:ChinaCandidate:A X XuFull Text:PDF
GTID:2370330590960639Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
As an open academic resource,academic conferences can disseminate scientific results in a clear,complete and timely manner,and promote communication among researchers.Therefore,academic conferences occupy an important position in the educational and scientific research field.However,in practical application environments,researchers often confined by the insufficient source of data low information matching rate,and inefficient resource management strategy when dealing with academic conference information.Thus it is hard to efficiently and accurately obtain information about interested conferences.Therefore,it is of great practical significance to study the processing technology of academic conference information and accomplish the personalized recommendation of academic conferences.Based on the requirements of extensive dissemination and efficient processing of academic conference information,this paper investigates the existing researches on solving academic conferences information processing problem in the engineering and academic field,analyzes the shortcomings in current research,and designs an academic conferences recommendation system based on the hybrid recommendation strategy.The main research contents are as follows:(1)The expansion of academic conference information sources.Our system gathers academic conference information mainly from two sources: emails and public conference sites.This article proposes an academic conference email filtering method based on keyword rules;On the other hand,this article applies webpage collection technology in the timed task to realize the real-time acquisition of the academic conference information from the website.(2)The extraction of key information in academic conference text.For mail information,this article proposes a method for the field recognition of academic conference text utilizing a text representation method combining TF-IDF and word vector,Moreover,we design the text pattern set to match the key information in the academic conference mail by the regular expression.For web page information,this article uses tag-based web page text parsing to extract key information from academic conference web pages.(3)The personalized recommendation of academic conferences.This article proposes a hybrid recommendation strategy for the personalized recommendation of the academic conference.The main idea is using content-based recommendation algorithm to calibrate and supplement the user-based collaborative filtering algorithm;In addition,the popularity of conferences divided by the domain are used as auxiliary information.(4)The management of academic conference resources.This article uses the interface call method to integrate third-party storage services to realize the systematic management of academic conference resources.Finally,this article collects real academic conferences datasets,and designs experiments to evaluate the performance of the three core schemes including the academic conference email filtering,academic conference text major field classification and academic conference recommendation,the results demonstrate the capability of the recommendation system for efficiently conference information processing.Simultaneously,this article designs functional tests by simulating the actual application scenarios,proving the functional integrity of the system.
Keywords/Search Tags:processing of academic conference information, recommendation system, hybrid recommendation algorithm, text classification
PDF Full Text Request
Related items